NODE LOCALIZATION AND ROUTING SCHEMES IN WIRELESS SENSOR NETWORKS

Abstract

Numerous applications of wireless sensor networks (WSNs) highly depend on the node location, such as maritime rescue, agriculture, and hazardous environments. GPS-enabled sensors are neither cost-effective nor energy-efficient. Finding the position of a target node (an unknown node) is known as node localization. Henceforth, the precise location of sensor nodes significantly impacts the performance of WSNs. Energy efficiency and network life are crucial concerns in WSNs due to the limited battery life of the sensor nodes. An efficient cluster head (CH)-based routing is a need of WSNs. Security is a major challenge in WSNs. Malicious nodes cannot be ignored in the hostile environment of WSN operations. Firstly, the thesis proposed an RSSI-based node localization. The localization error estimated using the RSSI and trilateration method is chosen as a fitness function for the Seagal Optimization Algorithm (SOA) for further minimization of the localization error. The SOA is modified using logistic chaotic maps and Lévy flights, known as C-SOA and LF-SOA. The simulation results illustrate that the performance of the LF-SOA is better than the C-SOA and SOA. Secondly, to improve the life of the sensor nodes, a cluster head selection (CHS)-based routing scheme is proposed. The Pelican Optimization Algorithm (POA) is modeled for an energy- efficient CHS and routing. The CHS is based on five fitness functions: energy of the sensor nodes, distance between cluster members (CMs) and cluster head (CHs), distance between CHs and base station (BS), node degree, and node centrality. These five fitness functions are applied to the POA to select the CHs. Further, the selected optimal CHs are participating in the routing process to send the aggregated data from the neighbor’s nodes (CMs). To choose the optimal path between CHs and BS (sink node), three fitness functions (residual energy of the CHs, distance between CHs and BS, and distance between CMs and CH) are chosen to model POA for optimal routing. The outcome of the proposed scheme (POA_Proposed) is energy-efficient, which enhances the network life. Lastly, a trust-aware node localization scheme (TANLS) is proposed. This work presents a novel model to resolve the shortcomings of existing localization techniques in WSNs. TANLS leverages trust-based approaches to accurately estimate the positions of sensor nodes while iv mitigating the impact of compromised or malevolent nodes on localization accuracy. The trust values of anchor nodes are evaluated against the reputation value of the anchor nodes. The highly trustworthy anchor nodes are subsequently selected as legitimate nodes for the localization process, and the unknown nodes get information from highly trustworthy anchor nodes to perform the NL. To enhance the performance of the localization, the Coati Optimization Algorithm (COA) is modeled using suitable fitness functions, and its performance is compared with the butterfly optimization algorithm (BOA) and particle swarm optimization (PSO). The simulations are performed using the COA_Proposed, BOA, and PSO. The comparative analysis among the COA_Proposed, BOA, and PSO demonstrates that the COA_Proposed outperforms the compared algorithms in terms of localization, localization computation time, and localization error.